The dog is not a wolf — origin, family and the myth of the pack

In a dog park an owner defends the muzzle-grip he has just used to clamp his dog’s snout shut. “That’s how the mother wolf does it with the cubs.” The other owner nods: “You have to take the alpha role.” If that sentence is true, every second dog-school trainer of the last fifty years is on the right track. If it isn’t, a large part of popular dog pedagogy stands on a 1947 zoo observation that its own author has since retracted.

The second is the case. Anyone who has read the relevant literature of the past three decades knows: the dog is not a wolf. He shares ancestors with the wolf — but he is not the same animal. And the family he lives in is not a pack. Four fields where the correction has been clearest.

Where the dog comes from

The genetics is strikingly clear. A 2015 study by Pontus Skoglund and colleagues, published in Current Biology, analysed the genome of a 35,000-year-old wolf excavated on the Siberian Taimyr peninsula. Recalibration of the mutation rate pushed the time at which the ancestors of today’s dogs split from the wolf line back: divergence happened before the Last Glacial Maximum, more than 27,000 years ago — not, as still often cited in the 2000s, at the human transition to settled agriculture.¹

A follow-up study from the Francis Crick Institute in London, published in Nature in June 2022, went further. Anders Bergström and colleagues analysed 72 ancient wolf genomes from Europe, Siberia and North America spanning the last 100,000 years and reached a finding that shifts the story again: the ancestors of all present- day dogs trace to an East Asian wolf population; dogs from the Near East and Africa additionally carry up to 50 percent ancestry from a second, south-west Eurasian wolf lineage — either via an independent domestication path or via later admixture with local wolves.² The dog is therefore not the domesticated descendant of one wolf but the product of a complex, at least two-stage interweaving of extinct wolf lineages.

In practical terms: the wolf from which the dog descends no longer exists. Today’s wolves are not the dog’s ancestors but distant relatives from a side branch — roughly the way your cousin is not your grandfather. Anyone reading dog behaviour off today’s wolves is asking the wrong relative.

Belyaev’s foxes — how domestication works

How a wary wild animal becomes a socially compatible housemate has been observable with unprecedented clarity since the late 1950s. Dmitry Belyaev, a Soviet geneticist at the Institute of Cytology and Genetics in Novosibirsk, began an experiment in 1959 that has since become inventory of behavioural biology. Belyaev selected silver foxes from a fur farm exclusively for one trait: tolerance toward humans. Each generation his team tested hundreds of pups; the tamest ten percent became the breeding base for the next generation.³

The results were more spectacular than expected. Within roughly ten generations the animals showed human-directed behaviour reminiscent of dog pups: actively seeking eye contact, tail wagging, attention to handlers. In parallel, morphological changes appeared that no one had selected for: floppy ears, curled tails, piebald coat patterns, reduced stress-hormone levels, longer reproductive seasons, juvenilised skull proportions. The literature collects this under the term domestication syndrome — a chain of accompanying traits that emerge with selection for tameness without being directly selected for.³

A 2019 re-analysis in Trends in Ecology & Evolution refined the picture: Belyaev’s source population came from a Canadian fur farm in which some pre-selection had already taken place; the evidence for some of the reported traits weakens when this is taken into account. The principle is not changed, but the reading sharpens: Domestication is not the flipping of a genetic switch, but a shift of whole distributions across many generations.⁴

Transferred to the dog: the journey from wolf to today’s bulldog, retriever or poodle took at least 27,000 years. And it has changed the animal itself, not only its behaviour: body plan, skull shape, hormonal balance and neural circuitry.

Schenkel in the zoo — how the alpha wolf was born

The most popular thesis on dog and hierarchy dates to an article from 1947. Rudolf Schenkel, a Swiss ethologist, observed at the Basel Zoo and other European zoos a group of wolves assembled from various fur farms and zoological gardens. What he saw he described in the terminology of chicken-pecking order: an alpha animal, a beta position, aggression around rank.⁵

For the assembled group, that description was not wrong. It was simply — and this is the decisive correction — not a description of a wolf pack in the wild. Schenkel’s animals were unrelated adults in close quarters, without genetic kinship and without freedom of choice. The human analogue would be a prison yard, not a family.

Schenkel’s observations were absorbed in 1970 by L. David Mech, an American wolf researcher, in his standard work The Wolf. The alpha terminology thereby became the lingua franca of wolf biology and — via dog training — the lingua franca of the trainer manual.⁶

Mech later spent 13 summers on Ellesmere Island, an Arctic island in Canada’s Northwest Territories (today Nunavut), observing free-living wolves. His correction, published in 1999 in the Canadian Journal of Zoology, was scientifically exemplary and personally courageous: what he had described as hierarchy was, in wild populations, simply family. The alpha pair consisted of the parents; beta and omega individuals were their offspring from the last two or three litters. What had looked like a pecking order was division of labour in pup rearing. Mech’s own words: »Dominance fights with other wolves are rare, if they exist at all. During my 13 summers where I observed the pack, I saw none.« The finding was nothing less than the scientific retirement of the alpha wolf from wild biology.⁶

Whether the trainer scene draws the consequence from this correction is a generational question. Anyone still speaking of alpha position, dominance training or rank order is doing so on a 78-year-old zoo observation that its own author no longer holds — and on the wrong animal.

What the family actually is

The second step: even if the wolf-pack hierarchy were real — which for wild wolves it isn’t — it would not apply to dogs. A 2002 Science paper by Brian Hare and colleagues at Harvard’s Department of Anthropology exposed dogs and wolves to the same test: a human pointed to one of two containers; only one held food. Dogs, even nine-week-old pups with little socialisation, followed the pointing reliably. Wolves — even hand-reared ones — did not, or only after long training.⁷

Hare’s conclusion: the cognitive ability to cooperate with humans is, in the dog, part of the domestication process. It is not learned — it is selected. What we feel from the dog as “family bond” is an evolutionary outcome, not a training achievement.

The Wolf Science Center group around Friederike Range in Austria has since refined the picture: hand-reared wolves can also read human signals if trained early and consistently. The difference lies not in the whether, but in the threshold: what the dog does on his own, the wolf learns only with explicit effort.⁸

For the question of what a dog-human family is, this has a practical consequence: it is a social bond, not a pack. Attachment research on the dog (Topál, Miklósi, ELTE Budapest) shows parallels with the parent-child bond in humans — not with leader-rank ordering. Whoever offers a dog trust, security and predictability gets cooperation. Whoever demonstrates dominance gets avoidance. Both are measurable.

Breeds — bred traits, with limits

The remaining popular question: what can I expect from the breed? Here the research situation is more complicated, because it has corrected itself in the last five years.

A widely cited study by Evan L. MacLean, Noah Snyder-Mackler, Bridgett vonHoldt and James Serpell, published in Proceedings of the Royal Society B on October 1, 2019, integrated behavioural data from over 14,000 dogs across 101 breeds with data on more than 100,000 genetic loci. For 14 behavioural traits a substantial share of the variance between breeds was heritable; trainability, predatory chasing, stranger-directed aggression and attention-seeking reached heritability values of 60 to 70 percent. 131 single-nucleotide polymorphisms (SNPs) could be linked to these traits, enriched in genes highly expressed in the brain.⁹

Three years later, in Science (April 2022), a study by Kathleen Morrill and colleagues from Elinor Karlsson’s group at the Broad Institute (Cambridge, Massachusetts) added a decisive nuance. On the basis of 18,385 owner surveys and genotypes of roughly 2,000 dogs, the authors found that there are indeed behavioural differences between breeds, but that individual variation within a breed is huge. Breed explains, on average, only about nine percent of the behaviour of a single dog.¹⁰ Both findings are compatible: breed says probability, not destiny.

In plain language, with the bulldog as example: an Old English Bulldog brings a different probability distribution than a Border Collie. A tendency toward tenacity, toward physical robustness, toward less autonomous working drive than in classic herding breeds. What the individual makes of this depends on socialisation, training, environment and simple personality. Anyone counting on “the breed is so” is wrong; anyone saying “the breed doesn’t matter” is also wrong.

What follows for owners

Four findings, sorted by trainer relevance:

1. Anyone measuring the dog against the wolf measures wrong. Today’s wolves are not the dog’s ancestors but distant relatives of a line that split 27,000 years ago. The genetics is unambiguous.
2. Dominance training rests on a 1947 zoo observation that its author retracted in 1999. Whoever still teaches it teaches the wrong edition.
3. The family is a social bond, not a pack. Dogs are evolutionarily selected for cooperation with humans; they give us for free what wolves only approximate under significant training. Trust, predictability and clear communication are the tools to put this evolutionary head start to use.
4. Breed says tendency, not identity. A bulldog is probably not the next German Shepherd substitute in service work, but neither is she a two-dimensional caricature of her breeding history. Whatever behaviour accrues comes from the individual animal — and from the owner.

Balance

The popular sentences about the dog — he is a domesticated wolf, you have to take the alpha role, the breed makes the dog — have all lost feathers, one after the other, in the research of the last thirty years. What remains is an animal that stands evolutionarily closer to us than we are willing to admit, but which is for that very reason its own being — not the domesticated wolf, but the result of 27,000 years of joint history with humans.

Anyone who has understood this stops subduing the dog and starts working with him. It is the older, quieter and far more effective school.

---